OHI British Columbia | OHI Science | Citation policy
There are two parts to creating this layer:
This process is completed entirely within this script.
Not applicable.
Raw data was provided by Woods Hole on December 16, 2014. This data is an update to the work done by Feely et al. (2009).
Raw data is read in for 1880 - 1889 and 2005 - 2014 datasets. Latitude, longitude, and aragonite saturation state \(\Omega\) are extracted for each year in each dataset. Land cells are set to NA.
Range of \(\Omega_{1880-1889}\) for global dataset: 0.073 - 4.961.
Range of \(\Omega_{2005-2014}\) for global dataset: 0.071 - 4.451.
The raw data, in WGS84 coordinate ref system, is rasterized, clipped down to a bounding box around the OHIBC region, and reprojected into BC Albers projection. This data is then resampled to 1 km resolution using nearest neighbor method.
The resulting raster layers (in rasterBrick format) include annual means for each coarse cell (approx 1 degree, from original resolution) for each year in the dataset. The two decades (1880-1889 and 2005-2014) are assigned to two separate rasterBricks.
Determine the mean cell value across the entire decade from 1880-1889, as a baseline for ocean aragonite saturation levels.
Baseline aragonite saturation state \(\Omega_{base}\) is based upon the decadal average from 1880-1889.
Deviation from aragonite saturation state is determined for each year in the study period:
\[\Delta \Omega_{year} = \frac{(\Omega_{base} - \Omega_{year})}{(\Omega_{base} - 1)}\]
Note that current is subtracted from the baseline; this way, a reduction in \(\Omega\) becomes a positive pressure value. It is then normalized by the current mean state; so a decrease in \(\Omega\) while the current state is high indicates less pressure than the same decrease when the current state is near 1.
\(\Delta \Omega_{year}\) is then modified to account for increases in aragonite saturation state (pressure = 0) and arag sat state less than 1 (pressure = 1).
Now that \(\Delta \Omega_{year}\) has been determined at a 1 degree cell scale, we interpolate that to the coastlines to fill NAs. Using raster::interpolate(), we select a random sample of points from the \(\Delta \Omega_{year}\) raster (for speed and memory) and apply the thin plate spline model (fields::Tps()).
Using 1000 m region ID raster to run zonal statistics on the pressures layers (within brick or separately?) we can calculate the regional average pressures for each year of study, and save as a .csv.
For each study year, calculate a running average of pressure score for each region, using the study year and the prior X number of years to determine the mean pressure.
| file_name | file_dir | filetype | uncomm_chgs | commit_hash |
|---|---|---|---|---|
| cesm_co2sys_1880-1889.nc | /Volumes/ohi/git-annex/globalprep/prs_oa/v2015/input | input | TRUE | NA |
| cesm_co2sys_2005-2014.nc | /Volumes/ohi/git-annex/globalprep/prs_oa/v2015/input | input | TRUE | NA |
| oa_rescaled_2005-2014_2000pts.tif | /Users/ohara/github/ohibc/prep/pressures/v2016/output/oa | input | FALSE | 6fac363 |
| ohibc_rgn_raster_1000m.tif | /Users/ohara/github/ohibc/prep/regions | input | FALSE | c474446 |
| ohibc_rgn_wgs84.shp | /Users/ohara/github/ohibc/prep/regions | input | FALSE | b014927 |
| ohibc_rgn.shp | /Users/ohara/github/ohibc/prep/regions | input | FALSE | b4848d1 |
| rbrick_a18_1km_raw.tif | /Users/ohara/github/ohibc/prep/pressures/v2016/int/oa | input | FALSE | 10da874 |
| rbrick_a20_1km_raw.tif | /Users/ohara/github/ohibc/prep/pressures/v2016/int/oa | input | FALSE | 10da874 |
| oa_rgn_pressures.csv | /Users/ohara/github/ohibc/prep/pressures/v2016/output/oa | output | FALSE | b687ba7 |
| pressures_oa_prep.Rmd | /Users/ohara/github/ohibc/prep/pressures/v2016/. | parent_script | TRUE | ec2b37a |
| pressures_lyr_fxns.R | /Users/ohara/github/ohibc/prep/pressures/v2016 | sourced_script | FALSE | 5f6534a |
| rast_tools.R | /Users/ohara/github/ohibc/src/R | sourced_script | FALSE | 61df74f |